Day 124: Too Many P’s and Not Enough Q’s
I just knew when I got back from the conference that my AP Statistics students didn’t have a handle on the two-sample methods for proportions. So much going on in these problems! When to use p-hat, p-naught, p-value. But my students didn’t even know what questions to ask.
Partner posters to the rescue! Of course, this was a 6am brainstorm idea for today’s activity….no pre-planning for me today! I decided I needed problems fast and remembered I had an old Practice of Statistics textbook lying around my room (actually it’s not that messy!). I found 15 practice problems using two proportions, but I didn’t want to be limited to either a confidence interval or significance test question. So I wrote up these directions:
Problem scenarios were randomly selected; partners were personally chosen. Lot’s of discussion and I was able to move about the room/groups to be sure the details of each process were clear in their minds.
One issue that kept coming up was why we use the hypothesized proportion to determine the standard deviation in tests, but use the p-hat when calculating the standard error in confidence intervals. Secondly, and related to the first issue, was why we pool for the test but not for the confidence interval. This activity provided an opportunity to check in with each group of 2-3 to be sure the concepts were solid.
I used an online stopwatch (the count down mode) to keep their efforts focused and moving along.
Once time was up, we posted using magnetic bulldog-clips from the Dollar Store. They are so handy, and I need more of them – note to self.
Once they were up, I handed out my trusty Inference Procedure Rubric so students could do a Gallery Walk and provide feedback to their peers. Didn’t quite finish the GW, probably because the posters were too close to each other; spread them out next time. All in all, kids walked out more confident in their understanding. Yay!